Are Global Temperatures Real?
The definition of ‘climate’ is often replaced by a proxy-value that doesn't exist anywhere other than in a computer..
‘‘Does reducing the ‘climate’ of the entire planet down to a single temperature value, expressed to two decimal places, make sense to you?’’
The New Climate Proxy
The meaning of the word ‘climate’ has changed.
No longer is it simply a description of the average weather.
‘Climate’ is now often reduced down to a single temperature value, expressed to extremely high levels of precision, and with no limit to the geographical area being considered.
Such values, often called something like 'The Global Temperature’ are more like a proxy for ‘climate’, rather than the climate itself.
Computer models are used not only to derive this proxy-value in the here-and-now, but also to extrapolate well into the past and far into the future, without losing any of it’s extraordinary precision.
And this number, as abstract as it may be, is taken VERY seriously! Many nations around the world are committing to economy-distorting ‘net-zero’ pledges in order to prevent something like a ‘1.5 degree global temperature rise’.
But these climate-proxy temperature values are not 'real'.
They don’t exist anywhere other than in a computer model.
Average temperatures - whether global or otherwise - do have some connection with reality by virtue of measurement data fed into the computer models used to derive them. But climate-proxy temperatures are, nonetheless, statistical constructs.
And even if you could measure everywhere, all the time, keep all the variables constant over time, and then determine some sort of mathematical average value (none of which are currently possible!) it still wouldn’t define ‘climate’ in a way that makes sense either conceptually or mathematically, as described below.
*you may also wish to refer to my piece from 2021, The Meaning of ‘Climate’ for a look at the definition of ‘climate’ throughout history.
Conceptual Problems
An Alien Thought Experiment
Imagine an alien came down to visit us here on Earth. During his time he managed to visit all of our continents.
On his way back home, the alien is asked to describe the Earth's climate.
He might say something like:
'It was warm in the tropics, sometimes reaching as high as 40 degrees. It's colder at the poles, where there is permanent ice and snow, with temperatures sometimes dropping as low as -50 degrees. The weather can be variable, especially in temperate zones where you can expect rain, snow, sunshine, depending on the season....'
…and so on.
The alien is talking about the average weather conditions over a given time in a particular location.
The description might not have a high level of precision, but it makes sense and it might even be useful (not least when deciding what clothes to wear!).
The new climate-proxy
Now let's pretend our alien visitor has some super-advanced form of computing at his disposal that allows him to actually determine an accurate, average temperature of the entire planet, taking into account it’s ever-changing (‘non-equilibrium’) and complicated nature.
If he were asked to tell us about the Earth's climate, and in response he whipped out his hand-held quantum computer, crunched some numbers, and told us:
'The Earth's climate is currently 26.56 degrees Celsius'
..it simply wouldn’t make sense!
(and it certainly wouldn’t help you plan what to wear!)
And yet that is effectively what we ARE being told whenever we hear ‘climate’ being reduced down to single temperature values.
Back down to Earth
Returning back to the real world, below is the top result I got when I asked Google what the Earth’s ‘climate’ was in 2023:
Does reducing the ‘climate’ of the entire planet down to a single temperature value, expressed to two decimal places, make sense to you?
That isn’t a ‘real’ number that exists anywhere other than in a computer.
It doesn’t make sense conceptually, and we can even demonstrate mathematically how it is not, in fact, a ‘real’ value.
A Mathematical Problem
Intensive and Extensive Variables
We can use good old thermodynamics to show how such a reductive way of expressing ‘climate’ doesn’t make sense mathematically. Bare with me, it’s not as scary as it sounds!
Let’s start with the question; ‘How would you calculate the Global Temperature?’
The obvious answer is something like: 'Calculate the average temperature measured across weather stations around the world.’
An 'average' is a value that we use to represent a range of values. The first step in coming up with such an average (lets assume the ‘mean’) is to sum a range of values, before dividing that sum by the number of values in that data set.
If you were to, for example, sum up people’s heights over a given population no conceptual problems will emerge. The sum of each individual’s height gives you a ‘real’ value that you could go out and physically measure (if the individuals were so kind as to stand on each other’s heads while you got your measuring stick out).
In thermodynamics this type of variable is called an 'Extensive Variable'.
But adding together other types of variables can create conceptual problems. An obvious example might be adding together currency exchange rates. The result doesn’t give you anything ‘real’.
Another example (perhaps familiar to my environmental science colleagues) would be when adding concentrations together… simply adding two pollutant concentrations together, for example, doesn't give you a 'total air pollution' value.
These are known as 'Intensive Variables'.
TEMPERATURE is an Intensive variable.
If you add two temperature values together, the answer IS NOT a temperature that you could go out and measure anywhere (other than by coincidence, of course), unlike the two temperature values that you started out with.
In other words, adding them together removes their underlying meaning.
To summarize, according to thermodynamics an average temperature calculated from a set of temperature data is NOT a ‘real’ temperature.
If you’re want to dig deeper into the question of whether the global temperature is actually ‘real’, I recommend this paper - Does a Global Temperature Exist?
Beware of the ‘climate-proxy’
Climatologists are therefore calculating, to extremely high levels of precision, a physical property (which they call ‘climate’), that doesn't actually exist.
Statistical constructs can, of course, be useful. But they can also be very misleading.
Statistics don’t lie, but liars do use statistics!
I think we can all agree that relying on statistics at least opens up the possibility of introducing bias, whether accidental or otherwise.
This is of course a big topic in it’s own right, but here I will quickly highlight some ways in which potential bias can (and often does) influence Global Temperature (aka ‘climate’) data.
Imagine every weather station on Earth takes a temperature reading each hour. At each station we find the average of all of those readings over a period of a year. Then you calculate the average of all those station-averages. Let’s call that ‘The Global Temperature’.
As explained previously, such a value isn’t a ‘real’ temperature, but nonetheless to track such a value might give us some idea as to general temperature trends over time, but ONLY IF certain conditions are met.
But to track a metric such as ‘Global Temperature’ in any meaningful way, it must be derived consistently over time.
Things like temperature instrumentation and measurement protocols need to be constant over the period in question - or if retrospective ‘adjustments’ are made to account for such changes, which they often are, yet more potential for bias is introduced.
The influence of human infrastructure is also a common problem for meteorologists, such as the ‘urban heat island effect’1 .
Studies that have removed the effects of such artificial heating from the raw data sets often claim that the temperature increase previously observed miraculously disappears!
There is also the question of how many stations are in the network over time, in which regions they and new ones are placed, and so on.
You can see that within climate science, keeping variables consistent over time to allow meaningful, long-term comparisons is a major challenge.
The broader point I wish to make here is how, when using statistically constructed values, especially when they originate from large and inconsistently-derived data sets, the opportunity for introducing bias (which to most of us will go completely unnoticed) is very much enhanced.
Should we even call it 'Climate Change'?
In summary:
To use a highly reductive, single temperature value to describe ‘climate’ makes no conceptual or mathematical sense.
In such instances, despite the apparent high levels of precision we are presented with, the derived value is not ‘real’ and does not actually exist anywhere other than in the computer used to derive it.
This does not entirely remove the usefulness of such values, but they are useful only if we understand a) that they are statistically derived, and b) what variables and assumptions are built into the statistical calculations.
'Mother nature changes the weather, statisticians change the climate'
And while they might technically be more accurate descriptions, names such as
'Statistically-Constructed Global Temperature Change', or
‘Climate-Proxy Change’
just don't have quite the same ring to them as the more hard-hitting
‘Climate Change’ or
‘Climate Emergency’!
Be careful out there!
-T
"Urban heat islands" occur when cities replace natural land cover with dense concentrations of pavement, buildings, and other surfaces that absorb and retain heat.